depth section, Oil & Gas Glossary

A depth section is a display of processed seismic reflection data in which the vertical axis is scaled in units of distance, metres or feet, rather than the two-way travel time used on a conventional time section. Because the raw seismic experiment measures only the time it takes for a sound pulse to travel from the source down to a subsurface interface and back to a receiver, every standard seismic image is born in the time domain. Converting that time image into a depth image requires knowledge of how fast sound travels through every layer of rock above each reflector, and this is where depth sections become both powerful and difficult: the conversion is only as accurate as the velocity model that drives it. There are two fundamentally different routes to a depth section. The first is direct time-to-depth conversion, in which an already migrated time section is stretched vertically using a velocity field, a comparatively cheap post-processing step suitable when velocities vary mostly with depth and not much laterally. The second is prestack depth migration, in which the imaging algorithm itself uses a depth-domain velocity model to position reflectors correctly, accounting for ray bending at each interface; this is the rigorous choice wherever velocities change rapidly from side to side, such as beneath the structural complexity of the Alberta Foothills or around salt and reef edges. The velocity model is built by integrating several independent sources of subsurface information: stacking and migration velocities measured from the seismic itself, sonic and check-shot data from boreholes, formation tops picked in wells, and geological constraints on layer geometry. A depth section matters because wells are drilled in depth, not in time. A geoscientist can interpret a structure on a time section and still mis-predict the depth at which a drill bit will encounter the target by tens or even hundreds of metres if lateral velocity variation has distorted the time image, a discrepancy that can mean the difference between landing a horizontal well inside a thin Montney or Duvernay target window and missing it entirely. Depth sections also remove the velocity "pull-up" and "push-down" artifacts that make time sections misleading, where a high-velocity layer such as a carbonate can make deeper reflectors appear structurally higher than they really are. Because the depth image ties directly to the units engineers and drillers use, depth sections are far easier for non-specialists, well planners, and reserves auditors to work with, and they form the backbone of the structural maps used to position pads, calculate gross rock volume, and book reserves in the Western Canadian Sedimentary Basin.

Key Takeaways

Vertical axis is distance, not time: A depth section plots reflectors against metres or feet rather than two-way travel time. This makes the image directly comparable to well logs, formation tops, and drilling targets, which are all measured in depth, removing the interpreter's need to mentally convert time into a drillable subsurface position.
Only as good as the velocity model: The conversion from time to depth depends entirely on an accurate three-dimensional velocity field built from stacking velocities, sonic and check-shot logs, well tops, and geological constraints. Errors in the velocity model translate directly into depth errors that can place a target tens to hundreds of metres off.
Two routes, different rigor: Direct depth conversion stretches a migrated time section and is adequate when velocity varies mainly with depth. Prestack depth migration uses a depth velocity model inside the imaging step and accounts for ray bending at interfaces, making it the correct choice where lateral velocity variation is strong, as in the Alberta Foothills.
Removes velocity distortions: Depth sections eliminate pull-up and push-down artifacts, where a fast layer such as a Nisku or Leduc carbonate makes deeper reflectors falsely appear structurally high on a time section. Correcting these distortions reveals the true structural shape needed for accurate trap and volume assessment.
Drives drilling and reserves decisions: Because depth sections read in the same units as the drill string, they underpin well planning, horizontal landing-point selection, gross rock volume calculation, and reserves bookings. A reliable depth section is what lets a WCSB operator commit millions of CAD to a multi-well pad with confidence in the target depth.

Building the Velocity Model

The velocity model is the engine of any depth section, and its construction blends seismic and well data. Stacking velocities derived during processing give a first-pass field, but these are processing parameters, not true earth velocities, so they are refined with reflection tomography that iteratively updates interval velocities to flatten common-image gathers. Borehole control sharpens the model dramatically: sonic logs supply interval velocities directly, check-shot and vertical seismic profile surveys tie travel time to measured depth at the wellbore, and formation tops anchor the depths of key interfaces. In the WCSB, operators with dense well control across mature Cardium or Viking fairways can build tightly constrained models, while frontier Duvernay or Montney step-outs lean more heavily on tomographic seismic velocities until the first wells provide ties.

Time Migration Versus Depth Migration

Time migration is mathematically valid only when velocity varies vertically, because it does not account for the ray bending that occurs when a wave crosses a dipping interface between layers of contrasting velocity. Where the subsurface has strong lateral velocity change, time migration mispositions reflectors and a simple stretch to depth inherits those errors. Depth migration solves the wave equation directly in the depth domain using the velocity model, correctly bending rays at interfaces and placing reflectors in their true spatial position. The cost is steep: prestack depth migration is computationally expensive and demands an accurate model, so WCSB operators reserve it for structurally complex plays such as the Foothills thrust belt, while gently layered plains stratigraphy is often well served by depth-converted time migration.

Fast Facts

A velocity error of just 2 percent in the overburden can shift a predicted target depth by more than 40 metres (about 130 feet) at 2,000 metres, easily enough to miss a 10-metre-thick horizontal landing window. This sensitivity is why operators in the Alberta Foothills, where steeply dipping thrust sheets stack fast carbonates over slow shales, invest in iterative prestack depth migration that can take weeks of compute time, whereas a flat-lying plains Cardium survey may be depth-converted in a single afternoon.

A depth section is the end product of seismic imaging that begins with a clean stack and is refined through migration. It depends on accurate noise removal earlier in the flow, including the tail mute that strips ground roll before stacking, since residual noise corrupts the velocity picks the conversion relies on. The reflections it displays are the same primaries protected from ground roll during processing, and the velocities that drive the conversion are anchored by borehole ties that connect the seismic image to real measured depth in the well.

Real-World WCSB Scenario: Landing a Foothills Well Near Rocky Mountain House

An operator targeting a tight gas sand within a thrust-faulted structure in the Alberta Foothills near Rocky Mountain House finds that the time section shows an apparent four-way closure, but a fast Nisku carbonate in the hanging wall is suspected of creating a velocity pull-up that exaggerates the structure. Rather than risk a CAD 9 to 12 million deep well on a possibly artificial high, the team commissions a prestack depth migration, building a velocity model from three offset wells, sonic logs, and reflection tomography over roughly six weeks of processing.

The resulting depth section shows the true closure is smaller and 60 metres deeper than the time image implied, and the crest is laterally offset by 250 metres. The well is repositioned and reaches the target within 8 metres of prognosis, a result that would have been a costly miss had drilling proceeded from the distorted time section alone.